| Brain Magnetic Induction Impedance Tomography(BMIT) is a kind of imaging technology for measuring the distribution of brain tissue's electrical conductivity by applying the detection of electrical magnetic field. The inducted current in brain tissues is generated by the influence of alternating magnetic filed. The perturbation of excitation field caused by inducted current generates phase deviation between excitation signal and detection signal, which is in direct proportion to brain tissue's electrical conductivity. Through the detection of phase deviation, brain tissue's electrical conductivity can be obtained. The image of a certain profile of brain tissue can be obtained by a series of phase detection around the profile and by some algorithms of image reconstruction. So, the phase deviation detection is of great importance in BMIT imaging system.Now, studies of phase detection mainly focused on the hardware circuits and the techniques of frequency downconvertion, which had many disadvantages, including designing complexity, low ability of anti-noise, reduplicate calibrations and low accuracy. While, phase detection method by software is rarely reported and has some advantages, including designing simplicity, high ability of anti-noise, high efficiency and the facility of integration with imaging algorithm.The research used software algorithms to improve BMIT phase detection system, aiming to realize a software system with high accuracy and low error. Based on the study of feasibility with 1MHz sine signal and the advantages of downconvertion technique, the paper implemented the generation and acquisition of dual channel 200KHz sine signal with data acquisition board based on Labview7.0. The signal processing algorithms, including FFT method, correlation method and classical method were also obtained by programming with Matlab7.0. The phase deviation output was arranged in three ranges, which held 1 degree, 0.1 degree, and 0.01 degree step size respectively. The obtained results with the accuracy of 0.1 degree were compared with those of lock-in amplifier SR844 by analyzing error and nonlinearity. Ultimately, a detection system composed of signal generation, acquisition and processing was obtained. Through feasibility experiments and simulation experiments based on programming with LabView7.0 and Matlab7.0, we can draw a conclusion that software phase detection method can be well used in MIT monitoring system and lay a foundation for future research. The main content is as follows:1. Analyzed different signal processing methods, selected a suitable one for feasibility experiments.2. Studied characteristics of data acquisition board, used Labview7.0 to implement dual channel sin signal generation and acquisition. 3. Programmed by Matlab7.0 to implement a series of processing algorithms and a whole detection system.4. Used lock-in amplifier SR844 for hardware phase detection and compared with software detection.The performance analysis was based on the calculation of error and linearity, which showed that FFT and Correlation measurements represented with linearity of -19.68 dB and -19.06 dB, error of -1.72dB and -2.76 respectively in the resolution of 0.1 degree. The digital phase detection provided a kind of more effective method for MIT system.
|
PDF Full Text Request